6.5. Comparisons and recommendations

The data support the view that, in symptomatic patients with hip prostheses, three-phase bone imaging should be done first, and if the arterial and soft-tissue phases are negative, no leukocyte scan is needed. Positive results may be confirmed using ^99mTc-leukocyte imaging with 24-hour images. Based on the present and previous observations, a practical approach to the diagnosis of hip prosthesis infections would include an initial screening of the patients with three-phase bone imaging, where a sensitivity of 100% and an accuracy of 80–90% can be achieved. Three-phase bone imaging is widely available and economical, and its radiation dose is low, especially when compared to ¹¹¹In-labelled leukocyte imaging. With this approach, however, there would be a need for additional examinations in many patients. If the supplementary method were extended ^99mTc-leukocyte imaging, 98% accuracy would be achieved, whereas ^99mTc-ciprofloxacin imaging reached an accuracy of 97%. Use of ^99mTc-leukocytes may, however, be less convenient due to the time-consuming blood handling with the associated risks of infection. Lack of ^99mTc-ciprofloxacin uptake in normal bone marrow, which may sometimes be a confusing factor in the interpretation of leukocyte images, can also be considered an advantage related to this tracer. In addition, ^99mTc-ciprofloxacin imaging is independent of the number and function of white blood cells and thus advantageous in patients with neutropaenia.

In symptomatic knee prosthesis patients, the arterial and soft-tissue phases of the three-phase bone imaging were almost always abnormal, and this method hence does not seem suitable for screening purposes. It would seem adequate to begin with ^99mTc-leukocyte imaging with 24-hour images and to supplement this technique with comparative bone-metabolic and/or bone marrow imaging in positive cases. ^99mTc-ciprofloxacin may also prove to be useful, although our patient series was too small to allow definitive conclusions.

In the future, other specific infection/inflammation tracers may also become available for studying patients with a suspicion of infection in their orthopaedic prostheses.